For over a century, strong fibers of regenerated cellulose have been produced by the viscose and cuprammonium processes. The latter process was first patented in 1890 and the viscose process two years later. In the viscose process cellulose is first steeped in a mercerizing strength caustic soda solution to form an alkali cellulose. This is reacted with carbon disulfide to form cellulose xanthate which is then dissolved in dilute caustic soda solution. After filtration and deaeration, the xanthate solution is extruded from submerged spinnerets into a regenerating bath of sulfuric acid, sodium sulfate, zinc sulfate, and glucose to form continuous filaments. The resulting so-called viscose rayon is presently used in textiles and was formerly widely used as reinforcing in rubber articles such as tires and drive belts.
Cellulose is also soluble in a solution of ammoniacal copper oxide. This property formed the basis for production of cuprammonium rayon. The cellulose solution is forced through submerged spinnerets into a solution of 5% caustic soda or dilute sulfuric acid to form the fibers. After decoppering and washing, the resulting fibers have great wet strength. Cuprammonium rayon is available in fibers of very low deniers and is used almost exclusively in textiles.
More recently other cellulose solvents have been explored. One such solvent is based on a solution of nitrogen tetroxide in dimethyl formamide. While much research was done, no commercial process has resulted for forming regenerated cellulose fibers using this solvent.
The usefulness of tertiary amine N-oxides as cellulose solvents has been known for a considerable time. Graenacher, in U.S. Pat. No. 2,179,181, discloses a group of amine oxide materials suitable as solvents. However, the inventor was only able to form solutions with low concentrations of cellulose and solvent recovery presented a major problem. Johnson, in U.S. Pat. No. 3,447,939, describes the use of anhydrous N-methylmorpholine-N-oxide (NMMO) and other amine N-oxides as solvents for cellulose and many other natural and synthetic polymers. Again the solutions were of relatively low solids content. In his later U.S. Pat. No. 3,508,941, Johnson proposed mixing in solution a wide variety of natural and synthetic polymers to form intimate blends with cellulose. A nonsolvent for cellulose such as dimethylsulfoxide was added to reduce dope viscosity. The polymer solution was spun directly into cold methanol but the resulting filaments were of relatively low strength.
However, beginning in 1979 a series of patents were issued to preparation of regenerated cellulose fibers using various amine oxides as solvents. In particular, N-methylmorpholine-N-oxide with about 12% water present proved to be a particularly useful solvent. The cellulose was dissolved in the solvent under heated conditions, usually in the range of 90.degree. C. to 130.degree. C., and extruded from a multiplicity of fine apertured spinnerets or dies into air or other nonprecipitating fluids, such as nitrogen. The filaments of cellulose dope are continuously mechanically drawn in accordance with a spin-stretch ratio in the range of about three to ten to cause molecular orientation. They are then led into a nonsolvent fluid, usually water, to regenerate the cellulose. Other regeneration solvents, such as lower aliphatic alcohols, have also been suggested. Examples of the process are detailed in McCorsley and McCorsley et al. U.S. Pat. Nos. 4,142,913; 4,144,080; 4,211,574; 4,246,221, and 4,416,698 and others. Jurkovic et al., in U.S. Pat. No. 5,252,284 and Michels et al., in U.S. Pat. No. 5,417,909 deal especially with the geometry of extrusion nozzles for spinning cellulose dissolved in NMMO. Brandner et al., in U.S. Pat. No. 4,426,228, is exemplary of a considerable number of patents that disclose the use of various compounds to act as stabilizers in order to prevent cellulose and/or solvent degradation in the heated NMMO solution. Franks et al., in U.S. Pat. Nos. 4,145,532 and 4,196,282, deal with the difficulties of dissolving cellulose in amine oxide solvents and of achieving higher concentrations of cellulose.
Cellulose textile fibers spun from NMMO solution are referred to as lyocell fibers. Lyocell is an accepted generic term for a fiber composed of cellulose precipitated from an organic solution in which no substitution of hydroxyl groups takes place and no chemical intermediates are formed. One lyocell product produced by Courtaulds, Ltd. is presently commercially available as Tencel.RTM. fiber. These fibers are available in 0.9-2.7 denier weights and heavier. Denier is the weight in grams of 9000 meters of a fiber. Because of their fineness, yarns made from them produce fabrics having extremely pleasing hands.
One limitation of the lyocell fibers made presently is due to their geometry. They are continuously mechanically drawn and typically have quite uniform, generally circular or oval cross sections, lack crimp as spun, and have relatively smooth, glossy surfaces. This makes them less than ideal as staple fibers since it is difficult to achieve uniform separation in the carding process and can result in non-uniform blending and uneven yarn. In part to correct the problem of straight fibers, man made staple fibers are almost always crimped in a secondary process prior to being chopped to length. Examples of crimping can be seen in U.S. Pat. Nos. 5,591,388 or 5,601,765 to Sellars et al. where the fiber tow is compressed in a stuffer box and heated with dry steam. It might also be noted that fibers having a continuously uniform cross section and glossy surface produce yarns tending to have a "plastic" appearance. Yarns made from thermoplastic polymers frequently must have delustering agents, such as titanium dioxide, added prior to spinning. Wilkes et al., in U.S. Pat. No. 5,458,835, teach the manufacture of viscose rayon fibers having cruciform and other cross sections. U.S. Pat. No. 5,417,909 to Michels et al. discloses the use of profiled spinnerets to produce lyocell fibers having non-circular cross sections but the present inventors are not aware of any commercial use of this method.
Two widely recognized problems of lyocell fabrics are caused by fibrillation of the fibers under conditions of wet abrasion, such as might result during laundering. Fibrillation tends to cause "pilling"; i.e., entanglement of fibrils into small relatively dense balls. Fibrillation is also responsible for a "frosted" appearance in dyed fabrics. Fibrillation is believed to be caused by the high orientation and apparent poor lateral cohesion within the fibers. There is an extensive technical and patent literature discussing the problem and proposed solutions. As examples, reference might be made to papers by Mortimer, S. A. and A. A. Peguy, Journal of Applied Polymer Science, 60:305-316 (1996) and Nicholai M., A. Nechwatal, and K. P. Mieck, Textile Research Journal, 66(9):575-580 (1996). The first authors attempt to deal with the problem was by modifying the temperature, relative humidity, gap length, and residence time in the air gap zone between extrusion and dissolution. Nicholai et al. suggest crosslinking the fiber but note that ". . . at the moment, technical implementation [of the various proposals] does not seem to be likely". A sampling of related United States Patents might include those to Taylor, U.S. Pat. Nos. 5,403,530, 5,520,869, 5,580,354, and 5,580,356; Urben, U.S. Pat. No. 5,562,739; and Weigel et al. U.S. Pat. No. 5,618,483. These patents mostly relate to treatment of the fibers with reactive materials to induce surface modification or crosslinking. Enzymatic treatment of yarns or fabrics is currently the preferred way of reducing problems caused by fibrillation. However, all of the treatments noted have disadvantages and increased costs. A fiber that is resistant to fibrillation would be a significant advantage.
Kaneko et al. in U.S. Pat. No. 3,833,438 teaches preparation of self bonded cellulose nonwoven materials made by the cuprammonium rayon process. Self bonded lyocell nonwoven webs have not been described to the best of the present inventors' knowledge.
Low denier fibers from synthetic polymers have been produced by a number of extrusion processes. Three of these are relevant to the present invention. One is generally termed "meltblowing". The molten polymers are extruded through a series of small diameter orifices into an air stream flowing generally parallel to the extruded fibers. This stretches the fibers as they cool. The stretching serves two purposes. It causes some degree of longitudinal molecular orientation and reduces the ultimate fiber diameter. A somewhat similar process is called "spunbonding" where the fiber is extruded into a tube and stretched by an air flow through the tube caused by a vacuum at the distal end. In general, spunbonded fibers are longer than meltblown fibers which usually come in discrete shorter lengths. The other process, termed "centrifugal spinning", differs in that the molten polymer is expelled from apertures in the sidewalls of a rapidly spinning drum. The fibers are stretched somewhat by air resistance as the drum rotates. However, there is not usually a strong air stream present as in meltblowing. All three processes may be used to make nonwoven fabric materials and all three processes do not employ methods which continuously mechanically draw the fibers. There is an extensive patent and general technical literature on the processes since they have been commercially important for many years. Exemplary patents to meltblowing are Weber et al., U.S. Pat. No. 3,959,421, and Milligan et al., U.S. Pat. No. 5,075,068. The Weber et al. patent uses a water spray in the gas stream to rapidly cool the fibers. A somewhat related process is described in PCT Publication WO 91/18682 which is directed to a method for coating paper by modified meltblowing. Coating materials suggested are aqueous liquids such as "an aqueous solution of starch, carboxy-methylcellulose, polyvinyl alcohol, latex, a suspension of bacterial cellulose, or any aqueous material, solution or emulsion". However, this process actually atomizes the extruded material rather than forms it into latent fibers. Zikeli et al., in U.S. Pat. Nos. 5,589,125 and 5,607,639, direct a stream of air transversely across strands of extruded lyocell dope as they leave the spinnerets. This air stream serves only to cool and does not act to stretch the filaments.
Centrifugal spinning is exemplified in U.S. Pat. Nos. 5,242,633 and 5,326,241 to Rook et al. Okada et al., in U.S. Pat. No. 4,440,700 describe a centrifugal spinning process for thermoplastic materials. As the material is ejected the fibers are caught on an annular form surrounding the spinning head and moved downward by a curtain of flowing cooling liquid. Included among the list of polymers suited to the process are polyvinyl alcohol and polyacrylonitrile. In the case of these two materials they are spun "wet"; i.e., in solution, and a "coagulation bath" is substituted for the curtain of cooling liquid.
With the exception of the Kaneko et al. patent noted above, processes analogous to meltblowing, spunbonding and centrifugal spinning have never been used with cellulosic materials since cellulose itself is basically infusible.
Extremely fine fibers, termed "microdenier fibers" generally are regarded as those having a denier of 1.0 or less. Meltblown fibers produced from various synthetic polymers, such as polypropylene, nylons, or polyesters are available with diameters as low as 0.4 .mu.m (approximately 0.001 denier). However, the strength or "tenacity" of most of these fibers tends to be low and their generally poor water absorbency is a negative factor when they are used in fabrics for clothing. Microdenier cellulose fibers, as low as 0.5 denier, have been produced before the present only by the viscose process.
The present process produces a new lyocell fiber that overcomes many of the limitations of the fibers produced from synthetic polymers, rayons, and the presently available lyocell fibers. It allows formation of fibers of low denier and with a distribution of deniers. At the same time, the surface of each fiber tends to be pebbled, as seen at high magnification, and the fibers have a cross section of varying shape and diameter along their length, have significant natural crimp, and are resistant to fibrillation under conditions of wet abrasion. All of these are desirable characteristics that are found in most natural fibers but are missing in lyocell fibers produced by processes employing continuous mechanical drawing means.